{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2025,9,29]],"date-time":"2025-09-29T08:19:20Z","timestamp":1759133960875,"version":"3.41.0"},"reference-count":50,"publisher":"MIT Press","issue":"8","content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Neural Computation"],"published-print":{"date-parts":[[2017,8]]},"abstract":"<jats:p>We propose a nonparametric procedure to achieve fast inference in generative graphical models when the number of latent states is very large. The approach is based on iterative latent variable preselection, where we alternate between learning a selection function to reveal the relevant latent variables and using this to obtain a compact approximation of the posterior distribution for EM. This can make inference possible where the number of possible latent states is, for example, exponential in the number of latent variables, whereas an exact approach would be computationally infeasible. We learn the selection function entirely from the observed data and current expectation-maximization state via gaussian process regression. This is in contrast to earlier approaches, where selection functions were manually designed for each problem setting. We show that our approach performs as well as these bespoke selection functions on a wide variety of inference problems. In particular, for the challenging case of a hierarchical model for object localization with occlusion, we achieve results that match a customized state-of-the-art selection method at a far lower computational cost.<\/jats:p>","DOI":"10.1162\/neco_a_00982","type":"journal-article","created":{"date-parts":[[2017,5,31]],"date-time":"2017-05-31T18:01:02Z","timestamp":1496253662000},"page":"2177-2202","source":"Crossref","is-referenced-by-count":8,"title":["GP-Select: Accelerating EM Using Adaptive Subspace Preselection"],"prefix":"10.1162","volume":"29","author":[{"given":"Jacquelyn A.","family":"Shelton","sequence":"first","affiliation":[{"name":"Technical University Berlin, 10587 Berlin, Germany"}]},{"given":"Jan","family":"Gasthaus","sequence":"additional","affiliation":[{"name":"University College London, Gatsby Unit, London W1T 4JG, U.K., and Amazon Development Center, 10178 Berlin, Germany"}]},{"given":"Zhenwen","family":"Dai","sequence":"additional","affiliation":[{"name":"University of Sheffield, Sheffield, South Yorkshire S10 2TN, U.K."}]},{"given":"J\u00f6rg","family":"L\u00fccke","sequence":"additional","affiliation":[{"name":"University of Oldenburg, 26129 Oldenburg, Germany"}]},{"given":"Arthur","family":"Gretton","sequence":"additional","affiliation":[{"name":"University College London, Gatsby Unit, London W1T 4JG, U.K."}]}],"member":"281","reference":[{"key":"B1","doi-asserted-by":"publisher","DOI":"10.1371\/journal.pcbi.1003062"},{"key":"B2","first-page":"161","volume-title":"Advances in neural information processing systems","volume":"20","author":"Bottou L.","year":"2008"},{"key":"B3","doi-asserted-by":"publisher","DOI":"10.1080\/01621459.1991.10475138"},{"journal-title":"GP-select demo on gaussian mixture models","year":"2016","author":"Dai Z.","key":"B4"},{"journal-title":"NIPS Workshop on Modern Non-Parametrics: Automating the Learning Pipeline","year":"2014","author":"Dai Z.","key":"B5"},{"key":"B6","first-page":"243","volume-title":"Advances in neural information processing systems","volume":"26","author":"Dai Z.","year":"2013"},{"key":"B7","first-page":"3338","volume-title":"Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition","author":"Dai Z.","year":"2012"},{"key":"B8","first-page":"2400","volume-title":"Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition","author":"Dai Z.","year":"2012"},{"key":"B9","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2014.2313126"},{"key":"B10","doi-asserted-by":"publisher","DOI":"10.1162\/neco.1995.7.5.889"},{"key":"B11","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1111\/j.2517-6161.1977.tb01600.x","volume":"39","author":"Dempster A. P.","year":"1977","journal-title":"Journal of the Royal Statistical Society B"},{"key":"B12","doi-asserted-by":"publisher","DOI":"10.1109\/TIP.2006.881969"},{"key":"B13","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-642-28551-6_26"},{"key":"B14","doi-asserted-by":"publisher","DOI":"10.1162\/15324430260185619"},{"key":"B15","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2012.273"},{"volume-title":"Bayesian optimization for likelihood-free inference of simulator-based statistical models","year":"2015","author":"Gutmann M. U.","key":"B16"},{"key":"B17","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-642-15995-4_56"},{"key":"B18","first-page":"2689","volume":"15","author":"Henniges M.","year":"2014","journal-title":"Journal of Machine Learning Research"},{"journal-title":"Proceedings of the Twenty-Ninth Conference on Uncertainty in Artificial Intelligence","year":"2013","author":"Hensman J.","key":"B19"},{"key":"B20","first-page":"2897","volume-title":"Advances in neural information processing systems","volume":"25","author":"Hensman J.","year":"2012"},{"key":"B21","doi-asserted-by":"publisher","DOI":"10.1126\/science.7761831"},{"key":"B22","first-page":"1782","author":"Kingma D. P.","year":"2014","journal-title":"Proceedings of the 31th International Conference on Machine Learning"},{"key":"B23","doi-asserted-by":"publisher","DOI":"10.1016\/S0893-6080(99)00049-0"},{"key":"B24","first-page":"609","volume-title":"Advances in neural information processing systems","volume":"15","author":"Lawrence N.","year":"2002"},{"key":"B25","first-page":"244","author":"Le Q.","year":"2013","journal-title":"Proceedings of the 30th International Conference on Machine Learning"},{"journal-title":"MNIST database of handwritten digits","author":"LeCun Y.","key":"B26"},{"key":"B27","first-page":"2855","volume":"11","author":"L\u00fccke J.","year":"2010","journal-title":"Journal of Machine Learning Research"},{"key":"B28","first-page":"1227","volume":"9","author":"L\u00fccke J.","year":"2008","journal-title":"Journal of Machine Learning Research"},{"key":"B29","doi-asserted-by":"crossref","first-page":"689","DOI":"10.1145\/1553374.1553463","author":"Mairal J.","year":"2009","journal-title":"Proceedings of the 26th Annual International Conference on Machine Learning"},{"key":"B30","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV.2009.5459452"},{"key":"B31","first-page":"593","author":"Meeds E.","year":"2014","journal-title":"Proceedings of the Thirtieth Conference on Uncertainty in Artificial Intelligence (UAI)"},{"key":"B32","first-page":"1791","author":"Mnih A.","year":"2014","journal-title":"Proceedings of the 31th International Conference on Machine Learning"},{"key":"B33","doi-asserted-by":"publisher","DOI":"10.1007\/978-94-011-5014-9_12"},{"volume-title":"Columbia Object Image Library (COIL-100)","year":"1996","author":"Nene S. A.","key":"B34"},{"key":"B35","doi-asserted-by":"publisher","DOI":"10.1016\/S0042-6989(97)00169-7"},{"key":"B36","doi-asserted-by":"publisher","DOI":"10.1145\/1273496.1273592"},{"key":"B37","doi-asserted-by":"publisher","DOI":"10.7551\/mitpress\/3206.001.0001"},{"key":"B38","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevLett.108.058301"},{"key":"B39","doi-asserted-by":"publisher","DOI":"10.1214\/ss\/1177012420"},{"key":"B40","first-page":"1209","volume-title":"Advances in neural information processing systems","volume":"17","author":"Schwaighofer A.","year":"2004"},{"key":"B41","first-page":"2653","volume":"15","author":"Sheikh A.-S.","year":"2014","journal-title":"Journal of Machine Learning Research"},{"key":"B42","first-page":"2618","volume-title":"Advances in neural information processing systems, 24","author":"Shelton J.","year":"2011"},{"key":"B43","first-page":"2285","volume-title":"Advances in neural information processing systems","volume":"25","author":"Shelton J.","year":"2012"},{"key":"B44","doi-asserted-by":"publisher","DOI":"10.1371\/journal.pone.0124088"},{"volume-title":"Scalable Bayesian optimization using deep neural networks","year":"2015","author":"Snoek J.","key":"B45"},{"key":"B46","first-page":"2510","volume":"24","author":"Titsias M.","year":"2011","journal-title":"Advances in neural information processing systems"},{"volume-title":"Accelerating ABC methods using gaussian processes","year":"2014","author":"Wilkinson R. D.","key":"B47"},{"key":"B48","doi-asserted-by":"publisher","DOI":"10.1016\/j.tics.2006.05.002"},{"journal-title":"Divide and conquer kernel ridge regression: A distributed algorithm with minimax optimal rates","year":"2014","author":"Zhang Y.","key":"B49"},{"key":"B50","doi-asserted-by":"publisher","DOI":"10.1371\/journal.pcbi.1002250"}],"container-title":["Neural Computation"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.mitpressjournals.org\/doi\/pdf\/10.1162\/neco_a_00982","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,6,18]],"date-time":"2025-06-18T21:53:05Z","timestamp":1750283585000},"score":1,"resource":{"primary":{"URL":"https:\/\/direct.mit.edu\/neco\/article\/29\/8\/2177-2202\/8298"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2017,8]]},"references-count":50,"journal-issue":{"issue":"8","published-print":{"date-parts":[[2017,8]]}},"alternative-id":["10.1162\/neco_a_00982"],"URL":"https:\/\/doi.org\/10.1162\/neco_a_00982","relation":{},"ISSN":["0899-7667","1530-888X"],"issn-type":[{"type":"print","value":"0899-7667"},{"type":"electronic","value":"1530-888X"}],"subject":[],"published":{"date-parts":[[2017,8]]}}}